Navid Ghafarzadegan Associate Professor of Engineering and Systems Engineering at the University of Virginia Tech and Hejir Rahmandad Associate Professor of Dynamic Systems at MIT, used a simulation model to estimate the number of coronavirus patients from the beginning of the outbreak in Iran (Feb 20) to March 20 (Persian New Year), totaling over 916,000 and more than 15,485 deaths. Referring to three factors, official policies, citizens’ cooperation and the impact of weather conditions, the researchers estimate that if the outbreak is not controlled in Iran until late June, in the best-case scenario, 1,600,000 people will be infected and the death toll would surpass 58,000.
The researchers noted that “this study develops and estimates a dynamic model of disease spread and adapts it to official and unofficial data on disease and mortality rates” saying “the difference between this model and the current ones is that the test percentage variables and the level of social interaction are endogenous and receive feedback from the spread of the disease.”
They emphasized that in this model, official statistics, unofficial statistics and actual statistics are separate and dependent variables. The calibration of this model, taking into account official and unofficial statistics, leads to estimation of test percentage and percentage of unofficial statistics coverage, therefore, the actual number of patients is estimated separately.”
The two researchers emphasized that the model presented was “capable of reconstructing existing formal and unofficial data to a desired extent by estimating test ratios.”
Left: Simulation of official stats: Left axis: Official accumulative number of infected (black) and recovered (green). Right axis: Official accumulative number of deaths in thousands.
The results of this research suggest that, for various reasons, such as “mild symptoms in the majority of patients, limited laboratory capacity and test kits, and delay in the testing process, a small percentage, about 5.2%, of all patients in Iran have been tested, and among the dead, the post-mortem test is done in about 26% of cases.
Gafarzadegan and Rahmandad’s article states that “simulating this model has led us to estimate 493,000 current cases [90% confidence interval: 271,000 to 810,000 [on March 20]”. This model estimates the accumulation of cases (including patients and healed cases), from the beginning of the coronavirus outbreak to March 20, is over 916,000 and with a 90% confidence interval: between 508,000 and 5.1 million.
The researchers also reported the death toll from the coronavirus in Iran is more than 15,485 people, which in 90% confidence interval the estimate would be between 8,400 and 25,800 deaths.
Dotted line: 90% confidence interval estimates
They wrote that, given the model’s constraints, and the complex nature of the problem, special consideration should be given to the 90% confidence interval in assessing these estimates, which indicates that the estimates are uncertain. However, they stated that these estimates show “the depth of the disease’s spread across the country, lack of adequate testing to detect cases despite the growth in the number of COVID-19 tests, and the need for effective policymaking both at improving testing processes and reducing social interaction.”
The researchers note that “according to estimates from this simulation, it seems that the decline in social interactions prior to Nowruz (March 20, 2020) has been somewhat effective in slowing the growth of the disease.”
They also said that “in the long-term, the citizens’ performance in keeping physical interactions down, is the primary concern”, they warned “according to model assumptions, people would be more likely to come back to normal activity in the society after morbidity and mortality rate decreases, which may cause another peak in the outbreak of the disease.”
The two researchers also noted that “trends depend on the policies of the authorities and the impact of atmospheric conditions and, according to the model’s estimates, by late June in best case scenario; rapid disease reduction with serious policymaker attention to reduced interactions and positive impact of climate”, 1,600,000 infected cases and more than 58,000 may die. And in worse scenarios “the death toll from the coronavirus in Iran may reach 103,000”.
Navid Ghaffarzadegan and Hajir Rahmandad, citing the study, emphasized that “to avoid the worst-case scenarios, social interactions must be kept low by serious focus of officials on this aspect.”
They warned that “an early announcement of victory may lead to the return of the disease, and therefore policymakers must be confident the outbreak has ended.”
The two researchers reminded us that the problems associated with the spread of coronavirus are not specific Iran and now many parts of the world are involved with this problem, stressing that the scope of coronavirus spread in Iran was intense, but “these estimates should not be compared to the incidence and mortality rates in other countries, because official statistics in some other countries, for similar reasons, are probably a low estimate of the actual number of patients and their deaths.”
Translation of this article.